Pernicious Retargeting and Remarketing Fraud

Pernicious Retargeting and Remarketing Fraud

Advertisers were led to believe that if someone visited their site, they "expressed interest." So it would be good to retarget them with ads to get them to buy stuff. That may be true in theory, but most of us experience retargeting as "those creepy ads that follow us around the internet even though we have already bought the item." Bombarding users with more ads doesn't necessarily make them buy more from you; and in some cases it may piss them off so much that it makes them hate your brand. But beyond the wanton overuse of retargeting, most advertisers are not aware that retargeting is a favorite playground for bots. See the examples below.

No alt text provided for this image
data from FouAnalytics


Retargeting fraud

Whether you are Advertiser 1, which uses a lot of retargeting, or Advertiser 2 which uses less, you will note the color coding under each -- lots of dark red (bots). In its simplest form, retargeting fraud happens when bots deliberately visit an advertiser's site first, and then go to cash-out sites to cause ads to load on those sites. That's how bots help cash-out sites make higher CPMs. Advertisers pay higher CPMs for retargeting because they assume those folks are more likely to convert, simply because they visited the site before. Challenge this assumption. And run a simple experiment -- turn off retargeting campaigns for a week and see if there's any change in sales, leads, business outcomes, etc. Oh, and you may find the clicks continue despite the campaign being turned off. That's because the bots are simply re-clicking the click through urls, without any ads having been run.

Since bots caused the ad to load, they also click on the ads. So the CTR ("higher click through rate") that you got from retargeting campaigns doesn't mean they were working better, it just means more bots clicked on the ads. Advertisers see what they want to see, and higher click rates justified their wanton use of retargeting. Further, even if you look at increase in sales numbers, the users were going to purchase anyway, so you didn't need to pay for the retargeted ads (or ads of any kind) in those cases. It would have been a better use of ad budget to generate awareness, among users who simply don't know you yet. That's a far larger percentage of the market. Even if a small percentage of these new folks convert, that can mean significant upside in numbers of sales.


Remarketing fraud

Remarketing fraud is even more pernicious. Targeting ads at users that had visited your site before is retargeting. Targeting ads at users that had purchased from you before is remarketing. Again, in theory, it may seem that targeting folks who had purchased before will get them to buy again and buy more. But, if I purchased the 4k TV already, no matter how many more ads I see, I won't buy another 4k TV, or headphones, or even milk. People simply can't drink 5 quarts of milk instead of 4 quarts, no matter how many ads they get bombarded with. And folks don't need a second pair of headphones when they only have one head.

If you've followed along so far, you're probably wondering how remarketing fraud happens, because bots would not have purchased from the advertiser before. Right. Remarketing fraud is double bad because all of it involves claiming credit for sales that would have happened anyway or had already happened. Attribution platforms and analytics are falsified to make it appear that the remarketing campaigns caused the sales. This form of fraud is so silly and done in broad daylight that if you spent a second thinking about it, you will realize you're being "taken to the cleaners." In the analytics reported to you, you will even see clicks from CTV campaigns. How do you click a CTV ad? Does it open a webpage? Where? And how do you buy something after that? (Note this is not the QR codes you scan with your phone and then visit the site with a browser on your smartphone.)

Phantom "clicks" are being falsely written into your Google Analytics based on algorithms that help the remarketing vendor create the appearance that the remarketing campaigns are working. In addition to these phantom clicks, the vendor can also use bot traffic to load the url, complete with utm_source= codes that attribute the click to the remarketing campaign. That is the red part of the row in the data grid above marked "remarketing." Keri Thomas , Performance Media Director at Iris Worldwide added, "this kind of mis-attribution can even be seen in the raw Google Analytics data. You'll see the user arrives at pages that would not normally be their first entry into the website (like the shopping cart). When we looked deeper at the full URL path, the remarketing vendor took credit for users who had arrived at the site from abandon shopping cart emails. There's no way the vendor could have driven the sale if the user entered from such an email."

Rethink retargeting. Rethink remarketing. Reduce your spend in retargeting and remarketing (and see if anything changes). And redouble your efforts to prevent getting ripped off.


Quoting Travis Lusk : "The reason...it makes them look good at their job.

Retargeting is incredibly good at taking credit for driving outcomes that would have otherwise been considered organic. Or taking credit for outcomes driven by less measurable (attributable) channels like TV.

This is not to say it drives zero performance, but it is certainly orders of magnitude smaller than what it claims.



Further reading: 618 other articles by me on analytics, ad fraud, and digital marketing

Fanny Marcoux

Ecommerce Analytics Consultant | Google Analytics, Google Tag Manager & Looker Studio since 2016 | Question E-commerce Newsletter | A very special coworking Podcast

1 年

I'd say the solution is simple, yet difficult: think. I'm usually not for ads, but I know they can be powerful in a few cases. In the case of remarketing/retargeting campaigns, it feels off to target people who only visited your website or with the product they just bought. Why not target visitors who downloaded a certain PDF with a product linked to the info presented in that PDF? Or a customer with the same product after the time it takes to use it? For example, with milk, a week after? That sounds more logical. Finally, there's the issue of analytics tools wrongly attributing sales (or whatever conversions you have set up). That sounds like a solid reason to change the tool.

Domenico T.

Senior Data Science-Marketing Professional

1 年

You know that part about the ad bots still clicking even after the campaign is over strikes a familiar chord...I've seen this for close to 15 years now. It makes sense that the click bot gets configured/initiated when the bot encounters the advertiser's ads. Bots are not carefully monitoring when the campaign starts and stops or is temporarily paused...they just keep going to keep the wheels turning. Meanwhile, site analytics tracking still shows "traffic"...thanks! ??

要查看或添加评论,请登录

社区洞察

其他会员也浏览了